arXiv Science⌕ Search

arXiv · 2611.18650

Random matrix discrepancy -- CLuP algorithm and parametric RDT approach

Abstract

We study information-theoretic and algorithmic aspects of random matrix discrepancy in proportional regime. The problem amounts to finding $κ=\min_{\x\in\mB^n}\|\sum_{i=1}^{n}\bar{A}_i\x_i\|_2$, where $\bar{A}_i\in\mR^{d\times d}$ is a collection of $n$ iid GOE matrices, $\mB^n$ are vertices of the unit cube in $\mR^n$, and $α=\lim_{n\rightarrow \infty}\frac{n}{d^2}$ is a proportionality constant. Strong progress is made in several directions: (i) We develop a Random duality theory (RDT) framework and utilize its a lifted variant to establish lower bounds on $κ$, which precisely match the bounds obtained via the first moment in [81]. (ii) To address analytical challenges, we propose a negative Hopfield model as a proxy for implementation of a fully lifted RDT. On the 2nd lifting level, this model fits remarkably well between the information-theoretic lower bounds from [81] and the algorithmic upper bounds from [38]. (iii) On the 3rd and 4th lifting levels, parametric RDT concepts predict a sizeable computational gap for lower $κ$ values in the proxy model. Extending proxy information-theoretic fitting to algorithmic suggests a similar computational gap for matrix discrepancy in low $κ$ regime, aligning with [73]. (iv) We design a Controlled Loosening Up (CLuP) algorithm for matrix discrepancy. Its performance strongly matches our analytical results -- closely approaching information-theoretic predictions in lower $α$ ranges (indicating a small or nonexistent computational gap) and following the proxy model's algorithmic predictions in higher $α$ ranges (indicating a sizeable computational gap). Both the information-theoretic and algorithmic predictions match the corresponding phenomenologies of the symmetric binary perceptron (SBP) vector analogue in the entire $α\in(0,\infty)$ range [1,2,14,31,51,92,111].

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Mihailo Stojnic. 2026-10-15. Random matrix discrepancy -- CLuP algorithm and parametric RDT approach. https://arxiv.science/abs/2611.18650

Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite arXiv Science by Indiciae: BibTeX · RIS.

KEEP EXPLORING

Related papers

Motif enrichment as a driver of scale-free behavior in rewired random regular graphs

We study the statistics of rewired random regular graphs (RRGs) in a mixed ensemble, where the average number of triangles is controlled by the fugacity $λ$, while the number of vertices and the vertex degree are fixed. This model exhibits a phase transition at critical fugacity $λ_{cr}$ from a triangle-poor phase (TPP), in which the number of triangles is independent of the system size, to a triangle-rich phase (TRP), in which the number of triangles scales linearly with the system size. We estimate $λ_{cr}$ by comparing the entropy of TPP with the energy of TRP. Above $λ_{cr}$, the RRG becomes a two-phase system in which dense clusters are connected by a sparse scale-free sub-network characterized by a degree distribution, $P(d) \sim d^{-γ}$, with $γ\approx 2$, independent of the size of the whole graph and its degree. We attribute this behavior to an "emergent preferential attachment" induced by triangle motifs, describe the mechanism underlying its formation, and derive the exponent $γ$ within a mean-field approach. Numerically the clusters and the inter-cluster subnetwork are identified through the Ollivier-Ricci curvature of the edges: it is positive inside clusters and negative on the links that form the inter-cluster subnetwork. We find that most inter-cluster triangles are isosceles, with the base lying inside one cluster and the apex belonging to the inter-cluster network. It is conjectured that a spanning inter-cluster scale-free network is stabilized against complete defragmentation into a disconnected set of cliques, by a sparse, structurally rigid "gas" of such apex-shared isosceles triads.

cond-mat.dis-nn↗

Graph neural networks and the energetic cavity method for combinatorial optimization

We study the use of graph neural networks (GNNs) for finding approximate ground states of Ising models. Efficiently finding these ground states is of broad significance because many combinatorial optimization problems can be formulated as an Ising model with the appropriate choice of couplings and fields. Exactly solving these problems is hard but there are many good heuristic methods. A lineage of these heuristics build from mean-field approximations: one approach uses the leading eigenvector of an appropriately defined matrix, another is the min-sum algorithm, also known as the energetic cavity method. Without modification, GNNs perform worse than both of these methods. We consider small modifications to the GNN to incorporate these heuristics and find that this considerably improves performance. While the modified approach is much stronger than other deep learning approaches, we still find that simulated annealing is generally as good for the same computational cost.

cond-mat.dis-nn↗

Contact-number hyperuniformity at glass close packing in infinite dimensions

We investigate contact-number fluctuations of harmonic spheres near glass close packing (GCP), the amorphous jamming endpoint reached by compressing an ideal glass. Using replica theory, we show that the Gardner-unstable region shrinks toward jamming as the spatial dimension increases, allowing GCP to be approached along a stable one-step replica symmetry breaking branch in the high-dimensional limit. By coupling an external field to the local contact number, we derive the variance of the total contact number and the wave-number-dependent structure factor directly from the microscopic interactions. At GCP, the structure factor vanishes quadratically at small scaled wave numbers, demonstrating contact-number hyperuniformity. Upon weak compression above GCP, the structure factor develops a low-wave-number plateau whose height is proportional to the excess coordination. These results provide an analytical description of contact-number hyperuniformity and its breakdown upon compression.

cond-mat.dis-nn↗